The research proposed in this application investigates retarded- nonretarded differences in the processing of semantic information. Specifically, these studies address the possibility that the well documented intelligence-related differences in semantic processing speed are mediated by corresponding differences in the allocation of attentional resources to semantic processing operations and their components. Two related lines of research are proposed. One series of studies examines retarded- nonretarded differences in the development of "automaticity" in cognitive processing and differences in the degree to which basic semantic processes require attentional resources for successful completion. The second series of studies assesses retarded- nonretarded differences in the ability to flexibly allocate cognitive resources across task components. Performance deficits exhibited by the mentally retarded may result from deficiencies in either or both of these aspects of resource allocation. It is expected that the proposed experiments will generate some of the first data available concerning the possible relationship between mental retardation and the effective utilization of cognitive processing resources.